Rationale
Texture-based radiomics analysis of lung computed tomography (CT) images has been
shown to predict chronic obstructive pulmonary disease (COPD) status using machine
learning models. However, various approaches are used and it is unclear which provides
the best performance.
Objectives
To compare the most commonly used feature selection and classification methods and
determine the optimal models for classifying COPD status in a mild, population-based
COPD cohort.
Materials and Methods
CT images from the multi-center Canadian Cohort Obstructive Lung Disease (CanCOLD)
study were pre-processed by resampling the image to a 1mm isotropic voxel volume,
segmenting the lung and removing the airways (VIDA Diagnostics Inc.), and applying
a threshold of -1000HU-to-0HU. A total of 95 texture features were then extracted
from each CT image. Combinations of 17 feature selection methods and 9 classifiers
were tested and evaluated. In addition, the role of data cleaning (outlier removal
and highly correlated feature removal) was evaluated. The area under the curve (AUC)
from the receiver operating characteristic curve was used to evaluate model performance.
Results
A total of 1204 participants were evaluated (n = 602 no COPD, n = 602 COPD). There
were no significant differences between the groups for female sex (no COPD = 46.3%;
COPD = 38.5%; p = 0.77), or body mass index (no COPD = 27.7 kg/m2; COPD = 27.4 kg/m2; p = 0.21). The highest AUC value for predicting COPD status (AUC = 0.78 [0.73, 0.84])
was obtained following data cleaning and feature selection using Elastic Net with
the Linear-SVM classifier.
Conclusion
In a population-based cohort, the optimal combination for radiomics-based prediction
of COPD status was Elastic Net as the feature selection method and Linear-SVM as the
classifier.
Key Words
Abbreviation:
COPD (chronic obstructive pulmonary disease), CT (computed tomography), QCT (quantitative CT), HU (Hounsfield units), LAA950 (low attenuation areas below -950HU), CanCOLD (Canadian Cohort of Obstructive Lung Disease), GOLD (Global Initiative for Chronic Obstructive Lung Disease), ATS (American Thoracic Society), FEV1 (forced expiratory volume in one second), FVC (forced vital capacity), HU15 (HU value corresponding to the 15th percentile on the frequency distribution curve), LAC (low attenuation cluster), TAC (total airway count), Pi-10 (estimated airway wall thickness for an idealized airway with an Internal Perimeter of 10 mm), NJC (normalized join count), SERA (Standardized Environment for Radiomics Analysis), IBSI (Image Biomarker Standardization Imitative), GLCM (gray level co-occurrence matrix), GLRLM (gray level run length matrix), GLSZM (gray level size zone matrix), GLDZM (gray level distance zone matrix), NGTDM (neighborhood gray tone difference matrix), NGLDM (neighboring gray level dependence matrix), ROC (receiver operating characteristic), AUC (area under the curve), CI (confidence interval), BMI (body mass index), PFT (pulmonary function test), SVM (support vector machine)To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to Academic RadiologyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- Systemic manifestations and comorbidities of COPD.Eur Respir J. 2009; 33: 1165-1185https://doi.org/10.1183/09031936.00128008
- Rapid fall in lung density following smoking cessation in COPD.J Chronic Obstr Pulm Dis. 2011; 8: 2-7https://doi.org/10.3109/15412555.2010.541306
- Quantifying the extent of emphysema: factors associated with radiologists’ estimations and quantitative indices of emphysema severity using the ECLIPSE cohort.Acad Radiol. 2011; 18: 661-671https://doi.org/10.1016/J.ACRA.2011.01.011
- Spatial dependence of ct emphysema in chronic obstructive pulmonary disease quantified by using join-count statistics.Radiology. 2021; 301: 702-709https://doi.org/10.1148/RADIOL.2021210198
- Computed tomography total airway count predicts progression to COPD in at-risk smokers.ERJ Open Res. 2021; 7: 00307-02021https://doi.org/10.1183/23120541.00307-2021
- Airway wall thickening on CT: relation to smoking status and severity of COPD.Respir Med. 2019; 146: 36-41https://doi.org/10.1016/J.RMED.2018.11.014
- Quantitative CT lung imaging and machine learning improves prediction of emergency room visits and hospitalizations in COPD.Acad Radiol. 2022; https://doi.org/10.1016/J.ACRA.2022.05.009
- Inter- and intra-software reproducibility of computed tomography lung density measurements.Med Phys. 2020; 47: 2962-2969https://doi.org/10.1002/MP.14130
- “Density mask”: an objective method to quantitate emphysema using computed tomography.Chest. 1988; 94: 782-787https://doi.org/10.1378/CHEST.94.4.782
Zwanenburg A, Vallières M, Abdalah M A, et al. The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping, Radiology, 295, 2020, 328–38.
- Comparison of feature selection methods and machine learning classifiers for radiomics analysis in glioma grading.IEEE Access. 2019; 7: 102010-102020https://doi.org/10.1109/ACCESS.2019.2928975
- Machine learning methods for quantitative radiomic biomarkers.Sci Rep. 2015; 5: 1-11https://doi.org/10.1038/srep13087
- Breast tumor characterization using [18F]FDG-PET/CT imaging combined with data preprocessing and radiomics.Cancers. 2021; 13: 1249https://doi.org/10.3390/CANCERS13061249
- A novel CT-based radiomics features analysis for identification and severity staging of COPD.Acad Radiol. 2022; 29: 663-673https://doi.org/10.1016/j.acra.2022.01.004
- Learning with many irrelevant features.AAAI. 1991; 91: 547-552
- Artificial intelligence selection of relevant features and examples in machine.Artif Intell. 1997; 97: 245-271
- Irrelevant features and the subset selection problem.in: Machine Learning Proceedings. Elsevier, 1994: 121-129https://doi.org/10.1016/B978-1-55860-335-6.50023-4
- The efficiency of logistic regression compared to normal discriminant analysis.J Am Stat Assoc. 1975; 70: 892-898https://doi.org/10.1080/01621459.1975.10480319
- Random decision forests.IEEE. 1995; 1: 278-282
- Support-vector networks.Mach Learn. 1995; 20: 273-297https://doi.org/10.1007/BF00994018
- An introduction to kernel and nearest-neighbor nonparametric regression.Am Stat. 1992; 46: 175-185https://doi.org/10.1080/00031305.1992.10475879
- Induction of selective Bayesian classifiers.in: Uncertainty Proceedings. 1994: 399-406 (Accessed March 21, 2022)
- Neural networks and physical systems with emergent collective computational abilities.Natl Acad Sci. 1982; 79: 2554-2558
- Canadian Cohort Obstructive Lung Disease (CanCOLD): fulfilling the need for longitudinal observational studies in COPD.J Chronic Obstr Pulm Dis. 2014; 11: 125-132https://doi.org/10.3109/15412555.2012.665520
- Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary.American journal of respiratory and critical care medicine. 2013; 187: 347-365
Dis ATS-ARR, 1987 undefined. Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma. ci.nii.ac.jp. https://ci.nii.ac.jp/naid/10005111964/. Accessed February 23, 2022.
- Impact of image pre-processing methods on computed tomography radiomics features in chronic obstructive pulmonary disease.Phys Med Biol. 2021; 66245015https://doi.org/10.1088/1361-6560/ac3eac
- Advanced statistical matrices for texture characterization: application to cell classification.IEEE Trans Biomed Eng. 2014; 61: 630-637https://doi.org/10.1109/TBME.2013.2284600
- Texture analysis using gray level run lengths.Comput Graph Image Process. 1975; 4: 172-179https://doi.org/10.1016/S0146-664X(75)80008-6
- Neighboring gray level dependence matrix for texture classification.Comput Vis Graph Image Process. 1983; 23: 341-352https://doi.org/10.1016/0734-189X(83)90032-4
- Texural features corresponding to texural properties.IEEE Trans Syst Man Cybern. 1989; 19: 1264-1274https://doi.org/10.1109/21.44046
- Outlier detection and removal improves accuracy of machine learning approach to multispectral burn diagnostic imaging.J Biomed Opt. 2015; 20121305https://doi.org/10.1117/1.JBO.20.12.121305
- A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models.J Clin Epidemiol. 2019; 110: 12-22https://doi.org/10.1016/J.JCLINEPI.2019.02.004
- COPD prevalence is increased in lung cancer, independent of age, sex and smoking history.Eur Respir J. 2009; 34: 380-386https://doi.org/10.1183/09031936.00144208
- Precision of quantitative computed tomography texture analysis using image filtering: a phantom study for scanner variability.Medicine (Baltimore). 2017; 96: e6993https://doi.org/10.1097/MD.0000000000006993
- A survey of feature selection and feature extraction techniques in machine learning.in: 2014 Science and Information Conference. 2014: 372-378https://doi.org/10.1109/SAI.2014.6918213
- Acute exacerbation of a chronic obstructive pulmonary disease prediction system using wearable device data, machine learning, and deep learning: development and cohort study.JMIR Mhealth Uhealth. 2021; 9: e22591https://doi.org/10.2196/22591
Article info
Publication history
Published online: August 11, 2022
Accepted:
July 17,
2022
Received in revised form:
July 15,
2022
Received:
June 23,
2022
Identification
Copyright
© 2022 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.
